Gliomas are uniformly fatal primary brain tumors, the diagnosis of which has been greatly impacted by improvements in medical imaging techniques over the last several decades. However, a significant gap remains between the obvious goal of more effective therapy and the present understanding of the dynamics of the tumor's proliferation and invasion in humans in vivo. That gap pivots on the concept that treatment of gliomas fails because of the diffuse dispersal of glioma cells throughout the neural axis even before diagnosis: the spatial and temporal evolution of which has been shown to be of quantitative and clinical importance as well as predicable with our current modeling methodology. Further, every imaging technique has a threshold of detection leaving much of the dispersed tumor invisible on imaging. The long-term objectives of this proposal are to provide new tools designed to quantify and predict the net proliferation and dispersal of glioma cells accurately enough to quantify and predict response to radiation therapy that are validated by and compared against information obtained through routine medical imaging of individual patients. The specific aims are to investigate the use of a spatio-temporal bio-mathematical model as a metric for glioma concentration, dispersal, response to radiation therapy, and location of post-treatment recurrence of individual gliomas in living patients in sufficient time to impact clinical decision making. This involves a gross but necessary assumption that medical imaging such as T1-weighted, gadolinium enhanced, T2-weighted MRI and PET imaging techniques directly correlate with disease distribution and biology. As the primary clinical window into disease progression, imaging techniques are used as benchmarks and metrics against which accuracy and success of model predictions are measured. Methods involve modern techniques and tools including, co-registration of clinical imaging, 3D radiation dose- distribution maps and the 4D patient-specific, model-simulated movie of the spatio-temporal growth and dispersal of each glioma. Comparisons are made between the model predicted invasion and therapy response patterns and that observed on follow-up imaging and, ultimately, autopsy. PUBLIC HEALTH RELEVANCE: The relevance of this proposal to public health lies in its applicability to any individual patient (and to the composition of any proposed group of "similar" patients) who has a primary brain tumor (glioma) and is being treated or is being considered for radiation therapy. Since disease progression and response to therapy are largely gauged by changes in current imaging techniques, there is an inherent limit to the clinical observation of a glioma to a "tip of the iceberg" view. Tools to predict and assess the dispersal (invasion) of gliomas cells throughout the brain in addition to the response to therapy which we cannot view on imaging is essential to the development of new and effective therapies for this uniformly fatal tumor. Specifically, as radiation therapy is targeted towards the dispersed glioma cells, peripheral to the imaging abnormality, it is necessary to calculate beyond the limits of imaging and to design mathematical models to dynamically assess that component of the tumor as well as take advantage of the tumor's proliferation rate in real time and in real patients.